Multiple linear regression + QQplots problem! Includes Pics!


I want to do multiple linear regression, but one of the requirements is the residuals to be normally distributed, and I can check that with QQplots but then the QQ plot shows it is about 95% of data fit into the normal line, but 5% is way off!

can I still proceed ?*or do I have to find a way to transform the data ?*

From what I understand of such things (not neccesarily a lot...) it looks like maybe you have a slight exponential curve... Which means that you should probably try fitting a quadratic and cubic term in a sequential regression as the last block. Looking at your output, I doubt that either will be significant, but it pays to be certain I guess.


Ninja say what!?!
Yeah...those model fit graphs look horrible. You might want to get someone to look at your model and data.


Ambassador to the humans
How many independent variables are you using? You might want to check the individual scatterplots with the dependent variable to see if you need higher order terms (or some other transformation).